The time cost of zooming itself must be counted
Aliases: navigation cost · multi-scale interaction
What it is
Zooming can reduce final pointing difficulty, but performing the gesture, waiting for animation, visually reorienting, and often panning afterwards carry zoom cost. Multi-scale pointing must be evaluated from task start through completion, not only by whether the enlarged target is easier to hit.
Why it happens
Before zooming, people decide whether it is worthwhile; after it, they relocate the target and interpret the scale change. When enlarged-target benefit is less than these fixed costs, direct pointing or a larger hot zone is faster. Cost matters most for short, low-precision work; on tiny or distant targets, width gain can offset it.
Studying it
Log zoom trigger, gesture duration, animation end, first reorientation, and final acquisition, then decompose total time. Compare direct operation and a larger-default-target condition across difficulty bands. Include novice canvas users because discovering and learning zoom strategy is itself a cost.
Where it stops holding
Shortcuts, wheels, automatic semantic zoom, and low latency can lower cost substantially; pinch input, remote rendering, and complex redraw can raise it. Need for global context also matters, so a millisecond estimate alone cannot decide.
Applying it
- Make zoom immediate and predictable, preserving focus so people do not lose the original target.
- Provide suitable default scale or local magnification for frequent precise work, avoiding repeated user deliberation.
- Measure end-to-end time and post-zoom disorientation; if zoom improves only the last click while slowing the whole task, revise defaults.